Enzymes have previously been combined with insoluble supports by using adsorption techniques [I. Langmuir and V. J. Schaefer, J. Am. Chem. Soc., 60, 1351 (1938)], but the products obtained suffered partial denaturation and the enzymes were progressively freed when they were in contact with substrates. Enzyme attachment was therefore not stable or showed poor resistance to external action.
Cellulose derivatives and enzymes have been combined by M. A. Mitz and L. J. Sumonaria [Nature, 189, 576 (1969)], who, for instance, obtained a carboxymethylcellulose azide from carboxymethylcellulose and then reacted this azide with a stabilized solution of an enzyme.
Stratis Avrameas (The Journal of Biological Chemistry, vol. 242, No. 7, pp 1651 to 1659, Apr. 10, 1967) reported copolymerization of human IgG and rabbit serum albumin at a pH between 4.5 and 5 with ethyl chloroformate to form a gel. As ethyl chloroformate was insoluble in the medium employed, the use of a non-aqueous solvent was necessary and the gel was that formed at the interface between two phases. Such gel could not in any way be formed into a self-supporting sheet or membrane; it is a polymeric precipitate with inadequate mechanical (strength) properties for such purpose.
A process is described in J. Epstein and B. Anfinsen's article, J. Biol. Chem., 237 (1962), dealing with coupling carboxymethylcellulose with ribonuclease or trypsin.
P. Bernfeld et al's article, Science, 142, 678 (1963), described a process for making antigens and enzymes insoluble by entrapping them in latticcs of synthetic polymer. The process consists of mechanically entrapping soluble macromolecular products in the lattice of a highly crosslinked Polymeric material by polymerizing some synthetic monomers in an aqueous solution in the presence of the biologically-active macromolecular substance to be embedded.
Goldman et al. (Biochemistry, vol. 7, No. 2, Feb. 1968, pages 486 to 500) disclose that active Papain-Collodion membranes may be formed by using bis-diazobenzidine-2,2'-disulfonic acid as cross-linking agent, but it could easily be shown that various other crosslinking agents will not give rise to an active papaincontaining membrane. Therefore, Goldman et al's disclosure is limited to the use of a specific carrier, a specific cross-linking agent and a specific enzyme. Moreover, Goldman et al clearly teach that bis-diazobenzidine derivatives inactivate the protein to a large extent since too much of it is required to insolubilize the protein.
Hornby et al [Biochem. J., Vol. 98, pages 420 to 424 (1966)] describe the preparation of ficin chemically attached to CM-celluloses using a method described by Mitz and Summaria; they describe neither the use of cross-linking agents nor their effect on the activity of the protein-active substance.
Moreover, several authors described the preparation of water-insoluble derivatives of enzymes:
(1) by chemical attachment of the enzyme to a reactive polymer Bar-Eli, A. & Katchalski, E., Nature, Lond., 188, 856 (1960) and J. Biol. Che., 238, 1690 (1963); U.S. Pat. No. 3,574,062; Cebra, J. J., et al, J. Biol. Chem., 236, 1720 (1961); Levin, Y., et al, Biochemistry, 3, 1905 (1964); Mitz, M.A.m & Summaria, L. J., Nature, Lond., 189, 576 (1961); Manecke, G., Pure appl. Chem., 4, 507 (1962); Habeeb, A.F.S.A., Archives of Biochemistry and Biophysics, 119, pages 264 to 268 (1967)];
(2) by physical adsorption of the enzyme to a charged polymer [Mitz, M.A., Science, 123, 1076 (1946); Mc Laren, A.D., & Estermann. E. F., Arch. Biochem. Biophys., 61. 158 (1956); Barnett, L., & Bull, H., Biochem. biophys. Acta, 36, 244 (1959); and Nikolaev, A. Y., & Mardashev, S.R., Biokhimiya, 26, 641 (1961); and Nikolaev, A. Y., Biokhimiya, 27, 843 (1962);
(3) by entrapping the enzyme in the insoluble matrix of a cross-linked polymer [Bernfield, P., & Wan, J., Scioence, 142, (1963)]; and
(4) by cross-linking of an enzyme by a bifunctional reagent Habeeb, cited above; and Quische, F. A., & Richards, F. M., Proc. Nat. Acad. Sci., 52, 833 (1964)], the latter consisting of linking molecules of a pure crystallized enzyme carboxypeptidase, the enzymatic activity being thereby greatly reduced.
These previously known processes have several drawbacks, including, interalia:
yields of immobilized active protein are low and, especially when a carrier iith covalent bonding is used, they are strictly dependent on the reactive sites present thereon, PA0 the active protein is not securely attached, and PA0 the active protein is denatured during attachment.